Stable carbocations. CLXXVII. Protonation in superacid solution of

Aug 27, 1974 - ported. In the first three examples (isomeric C10H10 hydrocarbons), exposure to superacid at —135° gave exclu- sively the bishomoaro...
0 downloads 0 Views 497KB Size
700 J. Org. Chem., Vol. 40,No. 6,1975

Paquette, Oku, Farnham, Olah, and Ling

Protonation in Superacid Solution of Selected Exo-Methylene Polyunsaturated Alicyclics. Monomethyl 1,3-Bishomotropyliurn, Bicyclo[3.2.2]nonatrienyl, and Norbornadienyl Cations1 Leo A. Paquette,” Masayoshi Oku, and William B. Farnham2 Evans Chemical Laboratories, The Ohio State University, Columbus, Ohio 43210

George A. Olah* and Gao Liang Department of Chemistry, Case Western Reserve University, Cleveland, Ohio 44106 Receiued August 27,1974

‘H NMR and l3C NMR spectral studies of the protonation of 9-methylenebarbaralane (7), B-methylenetricycl0[6.1.0.0~~~]nona-2,6-diene (8), 9-methylenebicyclo(4.2.l]nona-2,4,7-triene (9), 2-methylenebicyclo[3.2.2]nona3,6,8-triene (13), 7-methylenequadricyclane (18), and 7-methylenenorbornadiene (19) in FSO3H-SOzClF are reported. In the first three examples (isomeric CloHlo hydrocarbons), exposure to superacid a t -135’ gave exclusively the bishomoaromatic l-methylbicyclo[4.3.0]nonatrienylcation (10). In contrast, analogous treatment of 13 (a fourth CloHlo) a t -80° led cleanly to the stable 2-methylbicyclo[3.2.2]nonatrienyl cation (14), the first example of this heretofore elusive system, The stability and spectral features of 14 signify no obvious antibicycloaromatic character or other destabilization. With 18 and 19, direct conversion to the 7-methylnorbornadienyl cation was observed. Spectral data concerning the ground-state polarization of the exo-methylene polyolefins are also presented and discussed.

Current interest in the electronic requirements for longicyclic stabilization3 has prompted generation of several biand tricyclic carbocatioiis under solvolytic conditions or in superacid solvents. The most intensely studied member of this group, the 7-norbornadienyl cation ( I ) , is widely recognized to be highly ~ t a b i l i z e dIts . ~ distorted nature provides supportive evidence for the claim that homoaromatic interaction5 is the major contributing factor to its “bicycloaromatic” character.6 Various attempts to generate the bicyclo[3.2.2]nona-2,6,8-trienyl cation (2) have resulted instead in exclusive formation of the 9-barbaralyl cation (3).6y7The

t

6 +

1

N

2

@

7

8

9

n J

+

5

-

3

N

N

Results and Disctlssion Protonation of CloHl,~Isomers. When protonated in FSO3H-SO2ClF a t -135O, hydrocarbons 7,11913J4 8,I1J4and ggbJ1J5 underwent ready conversion to the l-methylbicyclo[4.3.0]nonatrienyl cation (lo), the 1H NMR spectral

+

initial claim that this facile rearrangement arises because of inherent “antibicycloaromatic” dest,abilization in 27a has been refuted in favor of its access to a low-energy route leading to energetically favored 3.7d The bicyclo[4.3.0]nonatrienyl (4) and bicyclo[4.3.l]deca-2,4,7-trienylcations (5) enjoy charge delocalization of the bishomotropylium 4 is the ion to which 3 rearranges a t temt ~ p e . ~ In g s fact, ~ peratures of -125’ and above.7e The bicyclo[4.2.l]nona2,4,7-trienyl cation (6) is likewise very prone to skeletal

4 N

fins under long-life conditions and wish to report the direct observation of the methyl-substituted carbocations. With but one exception, the 9-methyl-9-barbaralyl c a t i o n , 7 g ~no~ ~ substituted derivatives of these ions have previously been investigated.

6

isomerization with formation of dihydroindenyl cations,Y despite its predicted “bicycloaromatic” nature.3JO In recently published work,ll one of our groups demonstrated that electrophilic addition to exocyclic methylene precursors of such cationic intermediates was a serviceable probe of possible longicyclic interactions. We have now investigated t,he protonation of several such alicyclic polyole-

183 4 I/

N

‘,o

E

properties of which were identical with those previously reported by Win~tein.~g The I3C NMR spectrum of 10 was obtained at -90’ using the Fourier transform technique and the carbon shifts (in parts per million relative to capillary TMS)16 are shown alongside the structure. For comparison purposes, carbon shifts are also given for the mono(11) and 1,3-bishomotropylium ions (12), recorded under similar conditions.l7 Of the seven “basal” carbons in 10, four appear a t rather low field (139-143 ppm) and the remaining three in more shielded environments (118-125 ppm). These shifts reveal that extensive charge delocalization is present in 10 such that the bishomoaromatic formalism suitably represents its electronic structure. However, even after account is taken of the rehybridization necessary for homointeraction, it is seen that the pair of methine bridges serve to reduce the level of delocalization as com-

J . Org. Chern., Vot. 40, No. 6,1975 701

Protonation of Exo-Methylene Polyunsaturated Alicyclics pared to 11. In the latter case, the ring carbon shifts, exclusive of those due to C1 and C7, are more equally deshielded in accord with more equitable dispersal of the positive charge. The cyclic delocalization in 10 remains, however, a t a higher level than in 12, where the range of relevant carbon shifts is now greater than 40 ppm. That this ion possesses a ring current of lessened magnitude than that in 10 is also suggested by HMO calculations, which deduce that only 0.24 unit of positive charge is provided to the ethylenic moiety in 12 compared to 0.50 unit of positive charge donated away from the allylic part of 10 into its butadiene unit.7g Treatment of 2-methylenebicyclo[3.2.2]nona-3,6,8-triene (13)l18I4with FS03H-SO2ClF at -135' gave a dark brown solution which was stable a t -80' for days. l H NMR and 13C NMR spectral analysis of this solution signaled the formation of 14, the first example of the heretofore eluclass of cations. The charge ~ i v ebicyclo[3.2.2]nonatrienyl ~,~

-

13

distribution in 14, as reflected by the chemical shifts of the peripheral protons and ring carbon atoms, is totally unlike that in 10 and 12 and as such can only be viewed as a simple allylic ion. Reference ions having allylic part structures have previously been described.18 Of particular significance, the four olefinic protons on the two etheno bridges show normal olefinic character (Figure l),as do the attendant carbon atoms. Consequently, the spectroscopic properties of 14 are uniquely concordant with the absence of interaction between the allylic cation moiety and the double bonds. Furthermore, there is no evidence of inherent instability or unusual destabilization. Within narrow limits 14 gives every indication of being comparable to other allylic cations, notwithstanding the adverse inductive effect of the two additional double bonds. Cyclic charge delocalization involving one etheno bridge would make five of the carbon atoms carry a substantial part of the cationic character. However, this species would be electronically destabilized relative to the allylic moiety owing to its bishomocyclopentadienyl cation nature; as a result, it gains no importance. For similar reasons, antibicycloaromatic character is seen not to develop. The observation that 7 and 13 do not undergo protonation to give the same cation contrasts significantly with the earlier finding that tertiary alcohols 15 and 16 are both ionized directly to the 9-methyl-9-barbaralyl ion (17) a t -135'

5

L?

Is

under otherwise comparable conditions.7aJ2 Rather, proton transfer to 13, occurring as it does uniquely a t the exomethylene group, does not trigger homoallylic participation which would lead directly to 17. Is methyl substitution a t C2 of the bicyclo[3.2.2]nonatrienylcation as in 14 adequate to endow this cation with greater thermodynamic stability than its barbaralyl counterpart? This question is difficult to answer given only the present data. It would seem, however, that the direct conversion of 15 to 17 with resultant

-L,

PPM,H'

7

8

5

6

4

3

Figure 1. 'H NMR spectrum (60 MHz) of the 2-methylbicyclo[3.2.O]nonatrienyl cation (14). The six olefinic protons appear downfield while the two bridgehead protons and the methyl group are seen a t high field.

bypass of ion 14 arises by virtue of anchimeric assistance to C-0 bond cleavage in the protonated alcohol. Since bond breaking is not involved in protonation reactions of exomethylene compounds, energy requirements for cation formation are phenomenologically lessened and neighboring group participation is unnecessary. In principle therefore, protonation studies of exo-methylene derivatives are to be preferred to those involving tertiary methyl-substituted alcohols. Methylenequadricyclane and Methylenenorbornadiene in Superacid. When 1811,19 and 1S11J9were carefully added to FSORHS02ClF solutions a t both -120 and - 7 8 O , solutions of the 7-methylnorbornadienyl cation (20) were formed. The lH NMR spectrum is as expected for this

18

1 CH3

FS03H - S02CIF

/'

-120" 1.4

J

!2 ,/

"

1

species,2oshowing a methyl singlet a t 6 1.86 and multiplets for the bridgehead, olefinic, and homoaromatic protons a t 6 5.18, 6.32, and 7.70,respectively. 13C NMR data for the parent norbornadienyl cation have previously been reported (see 1).2l The 13C NMR parameters evidenced by 20 compare favorably and point up an interesting additional feature. The presence of the methyl group a t C7 serves to make the signals due to Cz and C3 almost identical in their chemical shift with thpse of Ca and CS. This probably arises from the expectedly22 lessened electronic demand a t C7, which in turn diminishes the need for homoaromatic stabilization and lessens conformational distortion. Ground-State Polarization of the Exo-Methylene Hydrocarbons. To infer longicyclic interaction in the neutral exo-methylene compounds would require that polarization of the exocyclic double bond operate in a direction which would electronically stabilize the particular molecule. In more classical systems such as fulvene (21) and heptafulvene (22),their differing reactivity characteristics have been ascribed to the existence of appreciable dipolar character which places opposite charge on the respective methylene groups. Access to cyclopentadienide and tropy-

702

J.Org. Chem., Vol. 40,No. 6, 1975

Paquette, Oku, Farnham, Olah, and Ling

lium ion character underlies the otherwise energetically unrewarding charge separation.

22 a

24 b

-105

2Ja

21 b ru

22a

ing up a small portion of its charge density to the remainder of the n system.

22b

N

N

The excessive shielding of the methylene carbon in 7methylenenorbornadiene (19)relative to 7-methylenequadricyclane (18) and 7-methylenenorbornane (23) has previously been commented upon.11J9 The appearance of the methylene proton signal in 19 (6 3.63) approximately 0.9 ppm to higher field than those in 18 and 23 is believed to point up the strong shielding arising from polarization which orients the negative terminus of the dipole away from the bicyclic framework. That structure 19b contributes significantly to the ground state is suggested further by its dipole moment (0.71 D) and 13C NMR parameters.lg In particular, the magnitude of A6 (>C=CH2) for 19 (99.5 ppm) is enormously larger than those for 18 (56.2 ppm) and 23 (61 ppm).

-

r 7 8 .I 177.6

wf 142

12a

19b N

,101.0

0-

97

,103

148 6

7 23 3

1176

-7 On this basis, the large A6 for 8 (100 ppm) would appear to be a direct consequence of an inherent tendency for development of ground-state 1,4-bishomotropylium ion character. Interestingly, the substitution pattern in 8b is not the same as that which is ultimately realized upon protonation. Experimental Section M a t e r i a l s . T h e preparation of t h e hydrocarbons has been previously described. P r o t o n N u c l e a r M a g n e t i c R e s o n a n c e Spectroscopy. lH NMR spectra were obtained using V a r i a n Associates M o d e l A56/ 60A a n d H A 100 NMR spectrometers equipped w i t h variable-temperature probes. E x t e r n a l tetramethylsilane (capillary) was used as reference. C a r b o n - 1 3 n u c l e a r m a g n e t i c r e s o n a n c e s p e c t r a were obtained using a V a r i a n Associates M o d e l XL-100 nmr spectrometer equipped w i t h a F o u r i e r t r a n s f o r m accessory, a s p i n decoupler, a n d a variable-temperature probe as previously described.24

Acknowledgment. Support of this work by the National Science Foundation is gratefully acknowledged.

23

1 8 hl

R e g i s t r y No.-7, 37816-60-7; 8, 38898-44-1; 9, 38898-39-4; 10, 53906-74-4; 13, 38898-40-7; 14, 53821-39-9; 18, 38898-42-9; 19, 37846-63-2; 20,19005-48-2.

N

References and Notes

The l3C NMR spectrum of 915 was seen to exhibit a chemical shift difference for the methylene group of 41 ppm, a value far smaller than that for 19 and actually distinctly less than the average A6 value of approximately 50-60 ppm seen for nonpolarized cyclic methylene compounds such as 23 and 25. This can be accounted for in terms of contributions from zwitterion 9b, an interpreta,107

9,o

+

9b N

tion which compares favorably with that offered in explanation of the 13C NMR data for 2-methylenebicyclo[3.2.1]octa-3,6-diene (24, A6 = 34.2 ppm, p = 0.59 D).23By way of comparison, 26 exhibits a quite normal A6 value of 47.8 ppm.23b Judging from the 13C NMR spectrum of 9-methylenebarbaralane (7),where a A6 of 38.1 is seen, it would appear that the exocyclic double bond is in this instance also giv-

(1)Considered as publication CLXXVll on Stable Carbocations from the Case Western Reserve laboratories.

(2) National Institutes of Health Postdoctoral Fellow, 1972-1974. (3)M. J. Goldstein and R. Hoffmann, J. Am. Chem. SOC., 93, 6193 (1971). (4)(a) S. Winstein and C. Ordronneau, J. Am. Chem. SOC., 82, 2084 (1960);(b) P. R. Story and M. Saunders, ibid., 82, 6199 (1960);84, 4876 (1962);(c) R. K. Lustgarten, M. Brookhart, and S. Winstein, ibid., 89, 6350 (1967);(d) M. Brookhart, R. K. Lustgarten, and S. Winstein, ibid., 89, 6352,6354 (1967):(e)R. K. Lustgarten, M. Brookhart, and S. Winstein, ibid., 90, 7364 (1968). (5) S. Winstein, 0.Rev., Chem. SOC., 23, 141 (1969);Chem. SOC.,Spec. Pub/., No. 21, 5 (1967). (6)J. B. Grutzner and S. Winstein, J. Am. Chem. SOC.,94, 2200 (1972). (7)(a) M. J. Goldstein and B. G. Odell, J. Am. Chem. SOC., 89, 6356 (1967); (b) J. C. Barborak, J. Daub, D. M. Follweiler, and P. v. R. Schleyer, ibid., 91, 7760 (1969);(c) J. C. Barborak and P. v. R. Schleyer, ibid.. 92, 3184 (1970);(d) J. B. Grutzner and S. Winstein, ibid., 92, 3186 (1970);(e) P. Ahlberg, D. L. Harris, and S. Winstein, ibid., 92, 4454 (1970);(f) J. S. Blair, J. Clark, and G. V. Meehan, Tetrahedron Lett., 3097 (1972);(9) P. Ahlberg, D. L. Harris, M. Roberts, P. Warner, P. Seidl, M. Sakai. D. Cook, A. Diaz, J. P. Dirlam, H. Hamberger, and S. Winstein, J. Am. Chem. Soc.. 94, 7063 (1972). (8)(a) D. Cook, A. Diaz, J. P. Dirlam, D. L. Harris, M. Sakai, S. Winstein, J. C. Barborak, and P. v. R. Schleyer, Tetrahedron Lett., 1405 (1971);(b) G. Schroder, U. Prange, N. S. Bowman. and J. F. M. Oth, ibid., 3251 (1970);(c) G. Schroder, U. Prange, and J. F. M. Oth, Chem. Ber., 105, 1854 (1972);(d) M. Roberts, H. Hamberger, and S. Winstein, J. Am. Chem. SOC., 92, 6346 (1970). (9)(a) A. S.Kende and T. L. Bogard, Tetrahedron Lett., 3383 (1967);(b) D. C. Sanders and H. Shechter, J. Am. Chem. SOC., 95, 6858 (1973);(c) A. F. Diaz, J. Fulcher, M. Sakai, and S. Winstein, ibid., 96, 1264 (1974). (IO)M. J. Goldstein, J. Am. Chem. SOC.,89, 6357 (1967).

J. Org. Chem., Vol. 40, No. 6, 1975 703

Reactions of N,N-Dimethylhydrazobenzenes in Acid Solutions L. A. Paquette and M. J. Broadhurst, J. Org. Chem., 38, 1893 (1973). (a) P. Ahlberg, D. L. Harris, and S. Winstein, J. Am. Chem. Soc., 92, 2146 (1970); (b) P. Ahlberg, J. E. Grutzner, D. L. Harris, and S. Winstein, ibid., 92, 3478 (1970). L. G. Greifenstein, J. B. Lambert, M. J. Broadhurst, and L. A. Paquette. J. Org. Chem., 38, 1210 (1973). 2. Goldschmidt and A. Worchel, Tetrahedron Lett.. 3621 (1973). M. T. Reek, R. W. Hoffmann. W. Schafer. and A. Schweig, Angew. Chem., h r . Ed. Engl.. 12.81 (1973). Starred values may be interchanged with those on the opposite side of the ring. L. A. Paquette, M. J. Broadhurst, P. Warner, G. A. Olah, and G. Ling, J. Am. Chem. Soc.. 95, 3386 (1973); G. A. Ohh. G. Ling, and J. S. Staral, ibid.. 96, 6233 (1974).

(18) G. A. Olah and G. Liang. J. Am. Chem. Soc., 94,6434 (1972). (19) R. W. Hoffmann, R. Schimler, W. S M f e r , and A. Schweig, Angew. Chem., ht. Ed. Engl., 11, 5 12 (1972). (20) G. A. Olah and A. M. White, J. Am. Chem. Soc.,91,6883 (1969). (21) G. A. Olah, G. Liang, G. D. Mateescu, and J. L. Riemenschmider. J. Am. Chem. Soc., 95,8898 (1973). (22) (a) P. G. Gassman and A. F. Fentiman. J. Am. Chem. Soc., 91, 1545 (1969); 92, 2549 (1970); (b) E. N. Peters and H. C. Brown, ibid., 95, 2397 (1973); (c) L. A. Paquette and I. R. Dunkin, IbM., submitted for publication. (23) (a) M. Sakai. J. Chem. Soc., Chem. Commun., 6 (1974); (b) R. K. Russell, R. E. Wingard, Jr., and L. A. Paquette, J. Am. Chem. SOC..98, 7483 (1974). (24) G. A. Olah and G. Liang, J. Am. Chem. Soc., 96, 189 (1974).

Benzidine Rearrangements. XIII. The Role of Reductive Scission. Reactions of N,N'-Dimethylhydrazobenzenes in Acid Solutionsl.2 Jiin-Duey Cheng3 and Henry J. Shine* Department of Chemistry, Texas Tech University, Lubbock, Texas 79409 Received September IO, 1974 N,N'-Dimethyl-p-hydrazoanisole (6b) in acetonitrile containing a small amount of hydrochloric acid underwent reductive scission to N-methyl-p-anisidine (7b, 70%). Other products formed were 2-amino-4-chloro-4',5-dimethoxy-N-methyldiphenylamine(the monodemethylated chloro-o-semidine, 10, 5%), 2,7-dimethoxyphenazine (9b) and 3-chloro-2,7-dimethoxyphenazine( 1 1) as a mixture which could not be separated quantitatively, and traces of 4,4'-dimethoxyazobenzene (12b).A radical was detected by ESR and is identified as the cation radical of 2,7-dimethoxy-5,10-dimethyl-5,1O-dihydrophenazine (18b). Anodic oxidation of 6b, its o-semidine (8b), and Nmethyl-p-anisidine each gave rise to the same ESR spectrum, also attributed to 18b.+. N,N'-Dimethyl-p-hydrazobiphenyl (6d) in acetonitrile-hydrochloric acid underwent reduction to N-methyl-p-aminobiphenyl(90%) along with other unidentified products, some of which contained chlorine. Molecular chlorine was not detected during reaction, and succinonitrile could not be found as a reaction product. A radical was detected during reaction under flow conditions only and its spectrum is attributed to 6d.+ rather than to the analogous dihydrophenazine (18d.+).N,N'-Diethyl-p-hydrazoanisole (16b)in acetonitrile-hydrochloric acid gave rise to an ESR signal attributed to 2,7-dimethoxy-5,10-diethyl-5,lO-dihydrophenazine (20b-+).In contrast the spectrum obtained with N,N'-diethyl-p- hydrazobiphenyl(16d) is attributed to 16d-+.Our results suggest that the major reducing agent in reductive scission of 6b is the first-formed rearrangement product, the o-semidine (8b),which is in turn oxidized successively to 18b, 9b, and 11. Analogous reactions are believed to occur with 6d. Oxidative demethylation of 6b and 6d also occurs. The results show, also, that the solvent is not a reducing agent in these reactions. The oxidizing agent is either the cation radical (e.g., 7b.+) formed by homolytic scission of the diprotonated hydrazo compound or the protonated hydrazo compound itself. In spite of the large effort that has been made by numerous research groups over the last few decades, the mechanism of the acid-catalyzed benzidine rearrangements remains unsolved. Of the several mechanisms that have been proposed, the polar transition state one from Ingold, Hughes, and Banthorpe remains the most attractive, although it is by no means universally a c ~ e p t e d . ~In- ~recent years, indeed, new proposals for the participation of ringprotonated rather than N-protonated intermediates have been although there is no direct experimental evidence in support of them as has been pointed out by BanthorpeYand by Shine.lo Apart from the problem of explaining the benzidine rearrangements there are the problems also of accounting for two reactions which accompany benzidine rearrangements. One of these reactions is disproportionation. This accompanies all acid-catalyzed benzidine rearrangements, sometimes to a very small degree and sometimes to an extent far larger than rearrangement itself.4 The mechanism of disproportionation is quite unk n ~ w n , ~ - ~ Jand " ~ "so is the way in which disproportionation and rearrangement may be linked together. Nevertheless, a large number of examples of disproportionation are known and the kinetic boundaries of the reaction have been established.6*11The second reaction which accompanies acid-catalyzed benzidine rearrangements is reductive scission, and this has been a much more fugitive and perplexing feature of the overall benzidine rearrangement

problem. It is this feature with which we are now concerned. Disproportionation of hydrazoaromatics leads to a 1:2 mole ratio of azoaromatic and scission amine (eq 1).Reduc2ArNHNHAr

1

-

ArN-NAr

2

+

2ArNH,

(1)

3

tive scission leads to 3 and is detected when the ratio 3:2 obtained from a hydrazoaromatic is more than eq 1 will accommodate. This was first noticed by Carlin and Wich in their very careful analysis of the products of reaction of p hydrazotoluene with acid,12 in which it was found that the ratio of the amounts of p - toluidine and p-azotoluene was a little larger than attainable by eq 1. Carlin and Wich concluded that some of their p - hydrazotoluene had undergone reductive scission, and assumed that solvent ethanol was the reducing agent. They did not, however, search for the oxidation product, acetaldehyde, because of the very small amount that would have been formed. Hammond and Clovis13 confirmed the finding12 that a small excess of p-toluidine was formed, and were also unsuccessful in accounting for the reaction, although they concluded that a transient oxidizing agent of unknown structure was probably associated with the high yield of p-toluidine. At the same time the reducing agent responsible for the excess of p - toluidine was not designated.